A novel representation in genetic programming for ensemble classification of human motions based on inertial signals

The use of sensing technologies and novel computational methods for automated motion detection can play a major role in improving the quality of life. Recently, researchers have become interested in employing the inertial sensor technology to record human motion signals as well as the new machine learning methods for signalbased motion detection. This manuscript proposes a novel method for human motion detection based on inertial sensors. The spatial information of a motion is first used in this method for geometric feature extraction. This manuscript also aims to introduce a novel ensemble learning approach through the genetic programing paradigm. To reduce the general complexity in the process of designing the proposed classifier, an initial population of binary trees (genes) is first created and then enhanced through genetic programing to select the best classifier. A complete experiment was conducted to evaluate the proposed ensemble classifier for the classification of inertial signals of human motions. According to the experimental results based on several well-known datasets of inertial signals, the proposed approach performed appropriately in comparison with the existing methods.

Automated summary

The manuscript introduces a novel method for human motion detection based on inertial sensors and a new ensemble learning approach through genetic programing. The method utilizes spatial information of human motion for feature extraction and enhances classifier optimization through genetic programing.